sync-skill
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires the agent to execute shell commands to manage directories and verify symbolic links using
ls -l. These operations target configuration folders for multiple AI agents, including.claude,.codex, and.hermes. - [INDIRECT_PROMPT_INJECTION]: The skill defines a process where user-provided inputs, such as the skill name, are used to construct file paths and execute commands. The absence of sanitization instructions for these inputs presents a risk if the input is maliciously crafted.
- Ingestion points: User-supplied skill names and the contents of the canonical skill directory.
- Boundary markers: No explicit delimiters or warnings to ignore instructions within the synchronized content are provided.
- Capability inventory: The skill utilizes directory creation, file listing (
ls -l), and symbolic link creation across different agent environments. - Sanitization: The instructions do not specify any validation or sanitization for the inputs used in file path construction or command execution.
Audit Metadata